the script exercise_influence_model.m generates the ego-level model data tables as described in detail in sections S2.2-S2.5 and S3.3 of the Supplementary Materials that will help us identify same day, one-day and two-day difference peer effects in running.

input data files: run_mat.mat
	   	  distance_mat.mat
	   	  duration_mat.mat
	   	  pace_mat.mat
	    	  TimeZone_mat.mat
	   	  StartTime_mat.mat 
	   	  PRECIPITATION_mat.mat
           	  TMAX_mat.mat
	   	  USERREL_USEDFOR_SOCIAL_INFLUENCE_wth_correlations.mat
	   	  App_Users_in_Graph_demographics.csv

output data file: ego_level_same_day_corr_thresh_0025_same_day.txt (only headers-redacted for legal reasons)
		  ego_level_same_day_corr_thresh_0025_one_day.txt (only headers-redacted for legal reasons)
 		  ego_level_same_day_corr_thresh_0025_two_day.txt (only headers-redacted for legal reasons)




the STATA script exercise_influence.do replicates the results presented in tables S4-S7 in supplementary materials and in figure 2 of the main text:

input data files: ego_level_same_day_corr_thresh_0025_same_day.txt (only headers-redacted for legal reasons)
	   	  ego_level_same_day_corr_thresh_0025_one_day.txt (only headers-redacted for legal reasons)
	    	  ego_level_same_day_corr_thresh_0025_two_day.txt (only headers-redacted for legal reasons)



